Top AI Product

Anthropic dropped something genuinely useful this week. [Claude Code Remote Control](https://code.claude.com/docs/en/remote-control) lets you start a coding session in your terminal, close your laptop, and keep steering the AI agent from your phone, tablet, or any browser. It’s not a gimmick — it solves a real annoyance that anyone who’s used Claude Code for longer tasks knows too well: you fire off a big refactor, and then you’re chained to your desk waiting for it to finish.

The clever part is that nothing actually moves to the cloud. Your local machine keeps doing all the heavy lifting — file system access, MCP servers, tools, project config, everything stays put. Your phone is just a window into the session. You scan a QR code from your terminal, it opens in the Claude app or browser, and you’re in. No SSH tunneling, no port forwarding, no fiddling with tmux. Anthropic built this as a direct response to the endless complaints in Claude Code threads about sessions getting stuck while developers were away from their desks.

The buzz around this has been pretty loud. It landed on [Product Hunt](https://www.producthunt.com/products/claude-code-remote-access) within days of launch and picked up solid traction. [MacStories did a hands-on](https://www.macstories.net/stories/hands-on-with-claude-code-remote-control/), [VentureBeat covered the release](https://venturebeat.com/orchestration/anthropic-just-released-a-mobile-version-of-claude-code-called-remote), [TechRadar wrote it up](https://www.techradar.com/pro/anthropic-reveals-remote-control-a-mobile-version-of-claude-code-to-keep-you-productive-on-the-move), and [DevOps.com](https://devops.com/claude-code-remote-control-keeps-your-agent-local-and-puts-it-in-your-pocket/) called it a pocket-sized agent. [Simon Willison tried it out](https://simonwillison.net/2026/Feb/25/claude-code-remote-control/) and was honest about the rough edges — he hit an “account not enabled” error on first try and noted it doesn’t support the `–dangerously-skip-permissions` flag, so you still have to approve every action manually. Fair enough for a research preview.

There are real limitations to know about. You can only run one remote session at a time, your terminal needs to stay open, and if your machine loses network for more than about ten minutes, the session times out. It’s also currently locked to Claude Max subscribers ($100-$200/month), though Pro access is supposedly coming soon. Given that Claude Code has reportedly hit a $2.5 billion annualized run rate with 29 million daily VS Code installs, opening this up to more users feels inevitable.

Is it perfect? No. Is it the kind of practical quality-of-life feature that makes you wonder why nobody built it sooner? Absolutely. Being able to check on a long-running agent task from the couch or approve a file change while grabbing coffee — that’s the kind of small freedom that actually changes how you work with these tools day to day.